Configural Frequency Analysis as a Searching Device for Possible Causal Relationships Die Konfigurationsfrequenzanalyse als ein Mittel des Erkennens möglicher kausaler Beziehungen

The empirical investigation of causal relationships typically requires knowledge about where cause - effect relationships might exist. If this knowledge is not available, researchers have no statistical methods at their disposal. This paper proposes using Configural Frequency Analysis (CFA) as a searching device for the identification of possible causal relationships. Three causal concepts - the Wedge, the Fork, and Reciprocal Causation - are considered. It is shown how certain patterns of CFA types and antitypes can be interpreted as conforming to these causal concepts. Data examples from nutritional, clinical, and developmental studies are used for illustration. Conceptual issues concerning the estimation of expected frequencies and the interpretation of CFA-identified causal relationships are discussed. This paper discusses Configural Frequency Analysis as a method for detecting possible causal relationships. For this purpose, causal relationships are defined at the level of categories of variables rather than the level of variables as such. The benefit from this approach is that various types of causal relationships can be specified, all of which can be detected using CFA. unknown publishedVersion